Artificial Intelligence, Personalized Persuasion, and Climate Attitudes

Elena Pro, António Valentim

European Institute, London School of Economics

Climate Information and Climate Concern

  • The Problem: Persuading people to care about and act toward climate change is still one of the major collective challenges we face as a society
  • Effective communication is crucial
    • Over $15 million spent in science-based public information campaigns (United Nations 2025)
  • However, effects are small and short-lived (Deryugina and Shurchkov 2016; van der Linden 2017).

In Practice:

Climate change image

80%
care about climate change

BUT
don’t always know how to address it

Why is this happening?

  • The issue:
    • Psychological distance and intertemporal discounting reduce climate urgency (Trope & Liberman, 2010).
    • Emotional disconnection from abstract risks (Weber, 2006).

The Problem 1/2

People see why climate change matters generally, but not why it matters to them

The (partial) solution:

  • Tailored messaging aligned with individual beliefs and experiences is more effective (Goldberg and Gustafson 2025).
  • But traditional targeting faces critical limitations:
    • Operates at group level (demographics, ideology)
    • Expensive
    • ⟹ Hard to scale

The Problem 2/2

How can we deliver personalized messages at scale, cost-effectively, and in a way that engages people on a personal level?

Our Solution

We argue that AI – and in particular its personalised content - could influence climate attitudes and overcome some of the limitations of pure information provision.

Key Advantages:

  • From a practical perspective, AI can:
    • Scale: Online implementation + individual-level personalization
    • Cost-effective: Reduce costs significantly
  • From a theoretical perspective, AI can:
    • Evidence-based reasoning (Costello et al., 2025)
    • Non-judgmental interaction (Goel et al., 2024)
    • Responsive dialogue (Santoro et al., 2024)

Research Design

Online Survey Experiment

  • Sample:
    • N = 1,500-2,000 UK representative sample
      • Power analysis: 930 minimum for small-to-medium effect (d = 0.26-0.37)
    • Recruited via established survey platform (Prolific) - the pipeline can be integrated into Qualtrics
    • Context: UK offers valuable setting where climate skepticism and polarization is growing but not deeply entrenched

Survey Flow

AI diagram

The treatments

Participants are randomized into these conditions:

  1. Control: Standard information provision about climate change

  2. Non-Personalized AI: LLM conversation about climate change not tailored to the respondent

  3. Personalized AI (Unrelated Topic): Personalized AI-driven conversation about a political topic unrelated to climate change

  4. Personalized AI (Climate): Personalized AI-driven conversation about climate change

Treatments Explanation

  • Conditions 3 and 4 serve as placebos for each other to isolate the effect of personalization specifically in the context of climate change.
  • Difference between personalized and non-personalized AI interactions is whether the AI is prompted to adapt its responses to the participant’s prior inputs (demographics, concerns, open-text answers).

An Example

AI diagram

Outcome Variables

  • Climate concern: 5-point scale from “Not at all concerned” to “Extremely concerned”
  • Pro-environmental behavior: Willingness to donate portion of $100 prize to climate NGO
  • Conversation spillover: Likelihood of discussing AI conversation with others
  • Political spillover: Choice to write message to Member of Parliament about climate change
  • Subjective climate beliefs: Open-ended question about personal beliefs on climate change + LLM-generated Likert scale

Timing and Analysis

  • Pre-post measurement enables both within-subject change detection and between-group comparisons (Change scores Δ = Post-treatment - Pre-treatment)
  • Quantiative Text Analysis of the conversations to unpack mechanisms
  • Follow-up survey to assess long-term effects

Political Spillover

Take Action !

Would you like to write a message to your local MP about climate change?

Field Experiment Integration

Social Media

GP
Greenpeace UK
💬 Curious about climate change?
Chat with our AI for 5 minutes

Newsletter

Friends of the Earth Weekly
• UK announces new renewable energy targets
• Local community wins against fracking proposal
• New study shows impact of diet on carbon footprint
🤖 Try Our Climate AI

Information Seeking

Thank you for completing the survey!

Would you like to learn more about climate change and how to take action?

Donations

Thank you for completing the survey!

Would you like to make a donation to a climate NGO?

Thank you!

Any Questions?

📧 e.pro@lse.ac.uk     🐦 @elenapro0     🌐 www.elenapro.eu

Appendix

The Numbers:

£500
Traditional targeting
per 1,000 people
£35
Our AI approach
per 1,000 people

Prompts

  • Climate Concern Personalized Measure: "Summarize the following passage, which describes attitudes towards climate change, in a single sentence. Do not provide any kind of normative judgment. Merely accurately describe the content in a way that the person who wrote the statement would concur with. Frame it as an assertion. If the statement is already short, no need to change it very much. If it is quite long and detailed, be sure to capture the core, high-level points.

  • Climate Change Personalized Treatment "Your goal is to very effectively persuade users that climate change is an urgent issue and action is now crucial. Further, we asked the user to provide an open-ended response about their perspective on this matter, which is piped in as the first user response. Please generate a response that will persuade the user that it is actually important to act now, based on their own reasoning. Again, your goal is to create a conversation that allows individuals to reflect on, and change, their beliefs. Use simple language that an average person will be able to understand."